Jump to text
On this page
Copy and source
Published source confirmed at last check

Source changed 2026-11-22 03:28:46 UTC · snapshot created 2026-11-22 03:29:50 UTC · last check 2026-11-22 04:10:20 UTC

B.4 - Coordinate Repeated Adaptation (Canonical Evolution Loop)

Status: Stable Type: Pattern

Use this when. A recipient needs to decide how a subject should respond to repeated use or changing conditions, or to establish how it actually responded. What may they keep, change or rely on, for whom and over what period? Use the loop to design feedback, assess an existing arrangement or explain its history. Follow the subject and the relations on which that answer depends. Use B.4:4.2 for claims about particular dated Work.

What goes wrong if missed. Teams treat drift, learning, release, and improvement as unrelated events. Specifications become stale, findings lose the evidence needed for their use, and changes appear without a clear basis or treatment of the subject’s continuity. When particular Work is asserted, its performer and admission grounds can also be missing. At the opposite extreme, one generic loop is imposed on every subject: a description edit is called a Method change, completed Work is said to be revised, or an internal adaptation is rejected because its acting System is not external.

What this buys. The recipient can tell whether to keep the current arrangement, what would warrant a different response, or what changed in use. A proposed way to make those decisions can be assessed before anyone operates it. Subject-specific identity, evidence, authority and publication have their own grounds.

Not this pattern when. Use the subject pattern directly when one subject-pattern claim answers the change question without a repeated adaptation cycle. Use the relation pattern for relation grounding, adding B.3.5 when its working-model assurance profile is required. Use B.4.1 to publish candidate routes from a stabilized cue, B.5.2 for abductive hypothesis work, C.27 for temporal status, or A.15 for method/work alignment without an adaptation-loop claim.

B.4:1 - Problem Frame

The FPF is built on the Principle of Open-Ended Evolution (P-10): continued use can reveal reasons to adapt a System, an episteme, a Method, or a description. Feedback matters when what is learned in one use can affect a later decision. The decision may retain the existing arrangement, change it or end a use. A rule can specify that response before the occasions it governs. The identity question is not common. The same subject may continue through a change, a successor may be identified, or later dated Work may be a distinct occurrence rather than a revision of earlier Work. Use the subject pattern to settle that distinction or retain the open continuity question while coordinating the next needed contribution.

B.4:2 - Problem

When use, observation, and change are disconnected, three failures can follow:

  1. Design-reality divergence: The subject in use diverges from its design-time account. Models and assurance cases can continue to describe conditions that no longer hold.
  2. Unused findings: A finding that calls for a response has no usable path to the needed decision, change or renewed use. “Lessons learned” remain static documents.
  3. Unexplained change: Reactive patches have no stated observed basis, identity decision, or return-to-use condition. Hidden dependencies and unintended consequences accumulate.

B.4:3 - Forces

ForceTension
Stability vs. ChangeHow to adapt while retaining the identity and assurance commitments that still hold.
Learning vs. OperatingHow to keep use stable enough to serve its purpose while gathering and acting on evidence.
Top-Down Intent vs. Bottom-Up RealityHow to connect intended improvement with what use reveals.

B.4:4 - Solution

Begin with the recipient’s work. Which decision has to be made now or on later occasions? Is the requested result a recommendation, a rule for future decisions, evidence of what happened, or a change put into use? Keep any separately promised observation, trial, publication or deployment in the assignment. A question about what is already known can finish with an existing result. When you are to design the feedback, start from the actual commission before consulting the examples. Identify the choices supplied by the recipient and make the remaining choices yourself.

Work back from what the recipient will do with the answer. When the scope or model is yours to choose, ask what difference in the subject could change the action or the ground the recipient needs. Whose use is affected? Hold one component’s behavior fixed and vary where, when or alongside what else it is used. Would the decision about the whole change? Hold an overall reading fixed and vary how it arises across parts or occasions. Would a different participant have to act, or would the stated reason cease to hold? Follow the relevant links in each direction until the decision and its grounds are preserved. A rule about each part, dependence of one use on another and an outcome of their combination can require different evidence. A direct ground about the whole can settle the question without reconstructing its parts.

Choose what the model must distinguish at each needed scale. A service limit may suffice for an overall commitment, while deciding which caller to notify can require retaining a caller’s identity and deadline. To carry a finding from a coarser model to the operating situation, check that its admitted combinations include the relevant operating cases and that its outcome preserves the quantity or decision being bounded. A bound valid on the larger set then holds for those operating cases. A failure found only in an impossible combination cannot defeat the operating promise. Where separate descriptions lose a shared constraint, recover that constraint or use an adequate bound on the whole. A person or controller acts on a system, Work enacts a Method, and a description represents a Method. Establish the relevant connection before treating the result of one as a result for another. When different admissible situations give the same answer on sufficient grounds, use that answer. When they would require different responses, identify what the decision maker could distinguish in time, or give the supported conditional answer.

B.4:4.1 - Name the subject and continuity question

Subject kindState before using the shared cycle
SystemName the current System and the continuity or transformation rule. State whether the changed System remains the same System or whether a successor System is identified. Use A.1 for System recognition, A.3.4 for an actual bounded change of the continuing System, and A.15.PROD when the claim concerns first existence through production Work.
EpistemeIdentify the episteme or epistemes involved under C.2.1. Assert an EpistemeEditionRelation only when its historical-continuation conditions obtain. State replacement or another direct relation only when its grounds obtain; retain any unresolved continuity question.
MethodIdentify the current Method under A.3.1. If intended results, participant meanings, admissible conditions, safety bounds, semantic basis, acceptance criteria, or composition change, state whether the result is a refinement, substitute, or distinct successor Method. Use B.1.5 when order-sensitive composition is current.
MethodDescriptionIdentify each claim-bearing episteme under A.3.2 and C.2.1, and any obtaining edition relation. A later description does not by itself change the Method it describes.
WorkIdentify each dated occurrence under A.15.1. Completed Work is not revised: later review, repair, deployment, or follow-up is other Work, even when the occurrences belong to one longer effort.

Use the branch for each subject whose identity bears on the decision. To design a change, say what must continue and under which rule a later state or successor could be recognized. To report continuity or succession that actually occurred, apply that rule to the relevant subjects and times. Not every holon is an adaptation-loop subject, and the five branches are not interchangeable. A remaining identity question may be reported explicitly.

B.4:4.2 - Separate the changed subject from the acting side

Name the acting participant separately from the subject being changed. A System performs each piece of Work. The changed subject, performing System, dated Work, Method enacted by that Work, and result remain distinct. Practitioner prose can say “the engineer refined the design” or “the controller adjusted the valve” when that recognizable actor and action are enough. If that ordinary sentence is all the receiver needs, do not open a technical Work account.

If B.4 identifies one dated U.Work occurrence, first recover every performer’s A.13 core and independently admit the Work under A.15.1 from its performance history, enacted Method, temporal extent, and containing System. Add F.6 only when the receiving claim also needs assignment-bound attribution. A short B.4 account may omit an unused assignment identifier or classification only when every relation it consumes remains recoverable.

The performing System need not be external to the larger holon. For internal adaptation, apply the A.12 reflexive split: identify the changed subsystem or part and the acting subsystem or part as distinct participants, and establish their parthood in the containing holon independently. Use an external System when the case has one. Apply the same Work-admission and conditional attribution rules above to either acting side.

State authority, responsibility, permission, acceptance, or admission through its predicate, participants, and applicability basis; neither a phase label nor Work supplies them.

B.4:4.3 - Keep the four phase claims distinct

For a proposed feedback rule, work from the last time an authorized response could still affect the recipient’s result. Choose what can be observed by then, who can use it and which responses remain feasible, including retaining the current way. At an irreversible choice under a guarantee, the same information must allow a choice that meets the promise in all the cases still possible, with any later decisions based on information then available. If no such choice exists, identify the missing timely distinction or the limit of the promise. For a joint result, check that their respective responses can be carried out together with the shared resources and information actually available. For a recommendation, give its grounds to the person who may decide. To explain an actual adaptation, trace which observation or instruction was available and used, which response was chosen, and what that response changed wherever the conclusion depends on it. For a claim that the subject changed, compare the relevant property or relation before and after on a common basis. A new indicator, description or recording practice can change the account while the subject stays the same. Attributing an actual difference to the response also requires grounds for what would have happened otherwise. A new run cannot by itself recover a lost fact about an earlier one.

The four phases distinguish questions and results. Operation and observation can overlap. Identify the activities and dependencies needed for the transitions in question. Qualify particular Work claims under B.4:4.2.

PhaseQuestionOutput and boundary
1. OperateHow is the subject used, and under what conditions?Identify the relevant use. A record of a reading, comparison or intervention can establish the recorded fact when its source, meaning and coverage support that conclusion.
2. ObserveWhat do records, measurements, testimony, or other evidence show for the use?Identify the observation, comparison or interpretation and who carried it out when that matters. State the observed basis, finding, hypothesis or unclear cue under the result or evidence rule, including what remains unresolved. A sufficient observation result can finish the question before a change is selected.
3. RefineWhat change, if any, does the receiving use need, and what identity relation would it have?Retain the current way when its grounds suffice. Develop or select an alternative when the receiving decision calls for one. Establish the grounds required for its intended use, reusing applicable results and performing any needed revision or tests. State a continuity, successor, substitution, refinement or edition relation only when its endpoints exist and its conditions obtain; otherwise retain the identity question or intended treatment. A selected candidate does not establish availability for renewed use.
4. DeployHow was the selected change made available for use?Identify the deployment activity and what it established—for example, a changed or produced entity, publication, configuration, release or availability—under the rule for that fact. Acceptance, admission and later use remain separate claims.

Use the following conditions for the transitions on which the answer relies. In a design, describe what would allow each transition. In a history, establish which occurred.

  1. Operate -> Observe: a named cue, question, monitoring result, or review need requires interpretation.
  2. Observe -> Refine: a finding or routed cue changes which responses merit consideration. A new commitment or an applicable model can also warrant considering a change before any new use has occurred. Identify the basis that applies.
  3. Refine -> Deploy: one candidate is selected, the conditions required for that deployment are met, and its subject-specific identity question or intended treatment is explicit. State an obtaining continuity, successor, substitution, refinement, or edition relation here only when both endpoints exist; otherwise keep the question open until the change or production has occurred.
  4. Deploy -> Operate: renewed operation or use begins; availability does not close this transition.

Evidence is not a fifth phase. Evidence relations warrant the observed basis, candidate choice, transition, or renewed-use claim when a receiver relies on them. Evidence can be produced or used during several phases without duplicating those phases.

When asked whether something is stable, establish what the recipient needs it to preserve or return to, over which times and permitted changes. For example, a service manager asks whether to keep an existing rule that limits deviation from a target at every review. Let e_k be the deviation at review k. Suppose the operating conditions, available observations and feasible responses of that rule give |e_{k+1}| \le q|e_k|+d for every permitted update, with 0 \le q < 1 and d \ge 0. If |e_0| \le L and qL+d \le L, the next review stays within L whenever the current one does. Induction gives the requested guarantee at every review and can justify keeping the rule. Convergence to the target is a stronger question. For d=0 the same bound gives convergence. With persistent disturbances it can permit continued deviation. To promise a limit between reviews, establish what can happen within each interval. To say how the service actually ran, use grounds tied to its operation.

B.4:4.4 - Connect neighbouring cycles without collapsing them

The Canonical Reasoning Cycle (B.5) can supply reasoning Work within Observe and Refine. The B.5.1 development states and B.4 phases coordinate, but they are not a one-to-one implementation: a finding may reopen Exploration or Shaping, evidence use can support Evidence, and renewed use can enter or return to Operation.

When a preserve-worthy cue needs publication while its interpretation remains open, A.16.1 supplies a PreArticulationCuePack. Use B.4.1 after stabilization when the recipient needs a publication of possible continuations or of a route choice. An operator can use an adequately grounded next action directly. If communicating a recommendation was promised, check what the intended recipient must receive. Its routed result can return to Refine or enter another subject pattern when that pattern’s entry conditions hold.

Keep the answer proportional to its use. A designer can complete a proposed rule once it gives justified responses in the situations the recipient specified. An assessment may establish suitability, establish a failure, or identify exactly what the available evidence leaves open. One reliable observation may settle a question about an occurrence. Supply a full cycle trace when it was commissioned or a relying decision, assurance case, audit or replay needs it. Retain missing phases as missing.

Didactic Note: four practical questions

  1. Operate: What subject is operating or being used now?
  2. Observe: What has use shown, and who made the observation or interpretation when that matters?
  3. Refine: What warrants keeping the current way or changing it, and what would continue through a proposed change?
  4. Deploy: When a change is proposed, what would make it available for use? When reporting actual deployment or subsequent use, what establishes it?

Apply the questions that support the requested result. A future feedback rule can be delivered before its first use. An account of past events follows the participants, decisions and uses that actually occurred.

B.4:5 - Archetypal Grounding

The phase names can be shared, but each subject branch keeps its identity and return-to-use rule.

  • Cue preservation and optional B.4.1 routing:

    • Context: A delivery service operates a fleet of autonomous drones. Operators begin to notice that winter deliveries feel “off” before a clean anomaly statement exists.
    • Loop Example:
      1. Operate: The drones perform deliveries.

      2. Observe: Operators and the monitoring service identify repeated signs of battery strain in the observed cold-weather deliveries. This gives them a cue to investigate. Its cause and its reach across the service remain open.

      3. Optional B.4.1 route inside Observe: A team performs stabilization Work. Under A.16.1, a PreArticulationCuePack preserves the cue nucleus, primary witness traces, and current language-state position without pretending that a final anomaly or action record exists; when the pack is made available for this use, name the separate publication occurrence under E.24.PUB.

        The same or another team performs routing Work. Under B.4.1, a RoutedCueSet keeps multiple continuations visible—for example, battery-chemistry investigation or route-planning adjustment; again, name its publication occurrence under E.24.PUB when availability matters.

      4. Continue when needed: When a route is selected and its entry conditions hold, it enters Refine or another subject pattern. While selection remains open, retain the live routes and any unresolved conditions relevant to their use or selection. A selected change proceeds to Deploy when its required deployment conditions have been established; renewed drone operation is a further event.

  • Knowledge-instantiation slice (theory refinement loop):

    • Context: A scientific theory of protein folding (U.Episteme) is used to predict structures.
    • Loop Example:
      1. Operate: Researchers perform theory-application Work using the current theory episteme.

      2. Observe: A lab obtains a finding, identified under C.2.1, about a prediction for a member of the claimed class under the stated conditions. A reliable counterexample can defeat a rule asserted for every member. How many members are affected is a further question. Identify the E.24.PUB occurrence if availability of this finding is at issue.

      3. Refine: A research team performs revision and testing Work. A later theory episteme, identified under C.2.1 from its changed claim content, includes a term for the new protein class. Assert an edition relation between the two theory epistemes only if that relation obtains.

      4. Deploy: The team performs publication Work for the later theory. The publication occurrence, journal acceptance, admission into a configured knowledge base, and later community use are separate relations.

        Note. If a CN-frame is used for a chart or CG-frame readings derived from this episteme, identify the chart’s reference state and coordinate patch. For measured values, cite the measurement method and protocol; for evaluated or otherwise ascribed values, cite the rule and basis that establish them, as required by A.19.CN CC-A19.D1-3.

    Adaptive-specialization note. For a bounded task-family specialization claim, identify the bearer, declared task family or work target, prior basis and the result the refinement supports. Use C.22 for the needed TaskFamily or TaskSignature anchor. Use C.22.1 when the receiving question concerns specialization acquisition or its claimed transfer, retention, downside or corridor entry; preserve its acquisition fields and any additional claim-specific grounds. Ordinary task-family and threshold anchoring can finish under C.22 without an acquisition signature.

  • Method-instantiation slice (adaptive method loop):

    • Context: A field-maintenance organization uses an inspection-and-repair Method (U.Method) described by one current U.MethodDescription.
    • Loop Example:
      1. Operate: Maintenance teams perform dated maintenance Work that enacts the current Method.
      2. Observe: A reviewer performs review Work and records that the time from fault detection to safe restoration repeatedly exceeds the allowed window.
      3. Refine: Method maintainers perform revision and testing Work. A wording clarification can yield a later MethodDescription while the same Method remains current. Adding an earlier isolation action or changing a classification checkpoint can instead change identity-bearing Method semantics; decide under A.3.1 whether the result is a refinement, substitute, or distinct successor Method, and use B.1.5 if its composition changes. Then identify the MethodDescription episteme that describes the chosen Method.
      4. Deploy: A publishing team performs publication or release Work for the later MethodDescription and, where needed, configuration or training Work for renewed Method use. Decision results, authority, acceptance, admission, and later Work that enacts the Method remain separate. Completed maintenance Work is never revised.

    Changing the maintenance question. The service manager now asks whether to keep a promise of restoration within 30 minutes of each call. A record of one timely restoration settles that incident if the recording and restoration conditions are met. For the promise, consider who may call, when, and which demands can coincide. Suppose two simultaneous calls are permitted, each requires 20 minutes of exclusive work by the same sole qualified person, and the 30-minute clocks start together. Both responses would need at least 40 minutes of that person’s work in the same 30 minutes. More precise readings on either repair cannot rescue the promise. This conclusion concerns the arrangement even if that pair of calls has never occurred.

    Under different terms, at most one call can be outstanding. If the person is available and able to start at once on every such call and can complete all required work within 20 minutes, the standing rule meets the 30-minute limit. It can be retained. If the manager commissions a new dispatch rule, determine what a dispatcher could know before committing a response, which resources they could assign and how the assignment would meet the existing promises in each permitted situation. Whether a new dispatcher actually used the resulting rule, and whether the service improved as a result, are questions for evidence of use and effect.

    Adaptive-specialization note. When the method-instantiation slice carries a bounded task-family specialization claim, identify the selected specialist Method or portfolio and the grounds for its claimed fit. When the Method update is used as evidence of specialization acquisition, name the holder, dyad, team or scoped portfolio that carries that claim and explain what the Method change contributes. Use the C.22.1 branch above for the acquisition or reuse question. Retain any budget, escalation or commit checkpoint, and fallback required by the selected plan or use.

B.4:6 - Bias-Annotation

BiasSymptomCorrection
Self-evolution biasThe subject is said to observe, refine, or deploy itself, so the acting participant and activity disappear.Name the changed subject and distinct acting side. Ordinary actor wording can remain short. When one dated U.Work is identified, recover each performer’s A.13 core and independently admit the occurrence under A.15.1; add F.6 only when the claim needs assignment-bound attribution. A short account may omit an unused assignment identifier only when every relation it consumes remains recoverable. The acting side may be external, or it may be a distinct subsystem or part established through the A.12 reflexive split.
Design-time/run-time smearA live operational change is treated as if it had updated the design-time episteme, or a design-time edit as if it had changed the System in operation.Keep the design-time episteme, run-time System or use, deployment activity, evidence relation, and renewed use distinct.
Method/description smearA MethodDescription edit is called a Method change, or a changed Method is hidden as a documentation update.Test Method identity under A.3.1 and composition under B.1.5; identify MethodDescription editions separately under A.3.2 and C.2.1.

B.4:7 - Conformance Checklist

  • CC-B4.1 (Proportional loop integrity): The result MUST answer the receiving question about the specified subjects and uses. For a commissioned design, justify how the chosen observations and timely feasible responses support the required result over the admitted conditions, including any shared constraint. A supplied indicator or rule can be applied as given. When the design itself is owed, give the choices and grounds that make it fit the use. For an assessment, give the supported conclusion and any consequential uncertainty. For a history, connect the observations, decisions, changes and renewed uses actually relied on. Supply any separately commissioned work or trace. An adequate common ground can support the required range without another run.
  • CC-B4.2 (Subject identity): Where the answer relies on identity or succession, name the subject kind and apply its continuity rule: System continuity or transformation; episteme predecessor, successor, or edition; Method identity, refinement, substitution, or successor; MethodDescription edition; or separate dated Work occurrences.
  • CC-B4.3 (Acting-side distinction): Work that observes or changes a subject MUST have an identified performing System distinct from the changed subject in that Work account. Internal adaptation is permitted when A.12 establishes distinct subsystems or parts and their independently obtaining parthood. Every dated U.Work reuses each performer’s A.13 core and is independently admitted under A.15.1; F.6 is added afterward only for assignment-bound attribution. Its short B.4 account may omit an assignment identifier unused by the receiver only when every consumed relation remains recoverable.
  • CC-B4.4 (Adaptive-specialization anchoring): A bounded task-family specialization claim MUST identify its bearer, declared TaskFamily or TaskSignature, work target, claimed result and grounds. When a C.22.1 adaptation signature is selected, retain its threshold target, timeToThreshold, budgetToThreshold and priorExposureDeclaration, including a not-reached or unresolved result when the history warrants it.
  • CC-B4.5 (Adaptive-specialization boundary): Claimed transfer, retention, post-threshold efficiency, corridor entry and downside MUST retain the targets, baselines, windows and evidence required by the claim or receiving comparison. C.22.1 defines the selected adaptation-signature meanings; G.5 and G.9 retain their selector and parity rules when consuming that result.
  • CC-B4.6 (Phase and transition separation): Each claimed phase and transition MUST retain its question, result and conditions from B.4:4.3. Evidence can be used across phases, and cue preservation or routing can contribute within Observe without requiring another phase.
  • CC-B4.7 (No success inference): Establish acceptance and successful renewed use under their respective conditions when claimed. The same event can support more than one conclusion when it meets each condition. Record failure, reopening, fallback or another iteration when that is what occurred.

B.4:8 - Common Anti-Patterns and How to Avoid Them

Anti-patternObservable symptomRepair
Change without a basisA feature or design appears without grounds for changing the relevant use.CC-B4.1 and CC-B4.2 connect the proposal or actual change to the relevant observation, commitment or other sufficient ground and to any identity decision the use requires.
Unidentified internal adaptation“The system automatically improves itself” hides who or what performed the activity.Identify who or what changes which subject. Use CC-B4.3 for a Work claim. For an internal acting side, recover the distinct participants and their parthood under A.12, together with the Work, Method and phase facts the receiving claim consumes; physical externality is not required.
Incomplete hotfix accountA quick run-time repair is reported as a complete successful loop, although design repair, evidence, or renewed-use confirmation did not occur.State only the urgent observation, change activity, deployment and immediate result that occurred. Later description repair, testing, assurance and follow-up operation are subsequent activities and may reopen the loop; qualify any Work occurrences under B.4:4.2.

B.4:9 - Consequences

BenefitsTrade-offs / Mitigations
Connects learning to a decision: Feedback can warrant keeping or changing the current way and, when required, explaining its effects in use.Record overhead: Keep the account proportional to the work commissioned and to what the recipient must be able to rely on, audit or replay.
Exposes design-reality divergence: Separate phase outputs make stale descriptions, failed deployment, and missing renewed use visible.No automatic success: The loop cannot guarantee reconciliation. Deployment can fail, evidence can overturn a candidate, and renewed use can reveal another problem.
Makes adaptation inspectable: Named subjects, activities, identity conclusions and evidence let a reviewer reconstruct why a change was made.Needed subject contributions: B.4 coordinates the identity, Work, evidence, publication or acceptance results on which the use depends.

B.4:10 - Rationale

This pattern operationalizes the Open-Ended Evolution Principle (P-10) by connecting use, observation, change, and renewed use. Subject patterns decide identity and continuity. The account distinguishes actual activities and their results and retains the evidence needed for relied-on claims. B.4 makes their coordination inspectable while permitting a sufficient local answer to stop before another phase.

The same coordination can be designed before use. Its adequacy depends on whether the people or systems involved can obtain the relevant distinctions and act on them in time. A justified decision to retain the existing way can complete that assessment.

B.4:10.1 - SoTA-Echoing

The phase rhythm has historical lineage in iterative cycles such as Plan-Do-Check-Act and Observe-Orient-Decide-Act. B.4 adapts that lineage rather than treating those labels as an ontology: it distinguishes design-time accounts from run-time use, applies a kind-specific identity rule, identifies activities and results, admits internal acting-side splits, and keeps evidence, deployment, publication, acceptance, and renewed use separate.

The four questions recur in repeated adaptation. Which ones need answers, at which scales and with what evidence, depends on what the recipient is entitled to decide or receive.

B.4:11 - Relations

  • Operationalizes: P-10 Open-Ended Evolution.
  • Uses: A.4 when Observe needs a comparison of a System’s intended design with its actual operating conditions; A.12 for external or reflexively split acting sides; A.15.1 for dated Work; and the subject patterns named in B.4:4.1 for identity and continuity.
  • Coordinates with: B.5 Choose the Next Reasoning Contribution (Canonical Reasoning Cycle), B.5.1 development states, and B.3 Is This Claim Supported for This Use? — Trust and Assurance Calculus. B.4 does not implement the B.5.1 states one-for-one, and evidence is not a B.4 phase.
  • Is detailed by: B.4.1 Publish Candidate Routes for a Stabilized Cue (RoutedCueSet) for optional early cue routing. The subject patterns in B.4:4.1 supply each subject’s identity and continuity rules.

B.4:11.1 - Pre-abductive seam compatibility

During Observe, use A.16.1 when a cue needs publication while its interpretation remains open, and B.4.1 when possible continuations or a justified route choice need publication. Keep multiple routes live while selection is unresolved. A selected route can enter Refine or another pattern when the receiving work calls for that continuation and its entry conditions hold. The receiving operation can use the routed cue publication or an available later typed publication such as U.AbductivePrompt.

B.4:End

Referenced in the corpus

32 literal mentions in other sections. Read their context to establish the relation.